Delay between a behavior and its reinforcer usually decreases the subjective value of the reinforcer, a phenomenon termed delay discounting. There is a well-established quantitative model, the hyperbolic discounting function, which describes delay discounting. Research with delay discounting has made substantial theoretical contributions to our understanding of drug abuse. For example, drug abusers discount the value of drugs more rapidly than they do monetary reinforcers, suggesting that discounting rate may depend on the nature of the reinforcer. However, experiments comparing discounting of drugs and money have studied a single magnitude of each reinforcer in a hypothetical choice situation. In humans, magnitude of a delayed reinforcer is a crucial determinant of discounting rate. Research with non-humans is beginning to contribute to our understanding of the effects of delay on drug choice. Recently, we found that monkeys discounted non-drug reinforcers more steeply than they did cocaine, a result opposite to that observed in drug abusers. This difference may be due to the nature of the specific reinforcers studied, or to the use of real reinforcers with monkeys vs. hypothetical reinforcers with humans. Alternatively, it may be a consequence of differences in the relative magnitudes of the reinforcers. We propose to investigate the effect of reinforcer magnitude and delay on the rate of discounting of drug and non-drug reinforcers in monkeys given allomorphic (i.e., drug/non-drug) choices. Aim 1 is to establish dose-response functions for cocaine when the choice is between an immediate injection of cocaine and the delayed presentation of food. Discounting functions will be established for different amounts of food available with various delays. This allomorphic choice situation (i.e., immediate drug vs. delayed food) is analogous to the one underlying discounting interpretations of drug abuse: drug abusers are assumed to impulsively choose immediate drug reinforcers over larger, delayed non-drug reinforcers. Such allomorphic choice has received little study to date. Aim 2 is to establish magnitude-response functions for food when the choice is between the presentation of an immediate amount of food and a delayed injection of cocaine. Discounting functions will be established for different doses of cocaine available with various delays. This allomorphic choice situation is important because drug abuse also can involve choosing between more immediate non-drug reinforcers and delayed drug reinforcers, as exemplified by the fact that drug abusers often devote considerable time and effort to procuring drugs. This situation has received virtually no discussion in the discounting literature. Our hypotheses for both Aims are that the discounting of the delayed reinforcer will be well described by the hyperbolic function and that rate of discounting will vary inversely with the amount of the delayed reinforcer. The results will have important implications for understanding the determinants of drug choice and, potentially, for the use of delay to reinforcement in a therapeutic context.